BACKGROUND: To compare response evaluation criteria in solid tumours (RECIST) and volumetric evaluation (VE) for colorectal cancer with liver-limited metastasis. PATIENTS AND METHODS: VE of liver metastases was performed by manual contouring before and after chemotherapy on 45 pairs of computed tomography (CT) images in 36 patients who suffered from metastatic colorectal cancer (mCRC) with liver metastasis only. Cohen kappa was used to compare the agreement between VE and RECIST. Pearson correlation was performed for their comparison after cubic root transformation of the aggregate tumor volumes. Logistic regression was done to identify clinical and radiographic factors to account for the difference which may be predictive in overall response (OR). RESULTS: There were 16 partial response (PR), 23 stable disease (SD) and 6 progressive disease (PD) cases with VE, and 14 PR, 23 SD and 8 PD with RECIST. VE demonstrated good agreement with RECIST (κ=0.779). Discordant objective responses were noted in 6 pairs of comparisons (13.3%). Pearson correlation also showed excellent correlation between VE and RECIST (r2=0.966, p<0.001). Subgroup analysis showed that VE was in slightly better agreement with RECIST for enlarging lesions than for shrinking lesions (r2=0.935 and r2=0.780 respectively). No factor was found predictive of the difference in OR between VE and RECIST. CONCLUSIONS: VE exhibited good agreement with RECIST. It might be more useful than RECIST in evaluation shrinking lesions in cases of numerous and conglomerate liver metastases.
BACKGROUND: To compare response evaluation criteria in solid tumours (RECIST) and volumetric evaluation (VE) for colorectal cancer with liver-limited metastasis. PATIENTS AND METHODS: VE of liver metastases was performed by manual contouring before and after chemotherapy on 45 pairs of computed tomography (CT) images in 36 patients who suffered from metastatic colorectal cancer (mCRC) with liver metastasis only. Cohen kappa was used to compare the agreement between VE and RECIST. Pearson correlation was performed for their comparison after cubic root transformation of the aggregate tumor volumes. Logistic regression was done to identify clinical and radiographic factors to account for the difference which may be predictive in overall response (OR). RESULTS: There were 16 partial response (PR), 23 stable disease (SD) and 6 progressive disease (PD) cases with VE, and 14 PR, 23 SD and 8 PD with RECIST. VE demonstrated good agreement with RECIST (κ=0.779). Discordant objective responses were noted in 6 pairs of comparisons (13.3%). Pearson correlation also showed excellent correlation between VE and RECIST (r2=0.966, p<0.001). Subgroup analysis showed that VE was in slightly better agreement with RECIST for enlarging lesions than for shrinking lesions (r2=0.935 and r2=0.780 respectively). No factor was found predictive of the difference in OR between VE and RECIST. CONCLUSIONS: VE exhibited good agreement with RECIST. It might be more useful than RECIST in evaluation shrinking lesions in cases of numerous and conglomerate liver metastases.